In recent years, with the development of technologies for processing data signals for various media such as sound, image, etc., it has become desirable to develop technology to produce multimedia representations without difficulty.
Hereinafter, media is defined as each elementary means to carry sound or images, for example, which when combined form a multimedia representation.
One prior art method for producing multimedia representations was to combine a plurality of media such as sound, image, etc. using an editor. An editing person would edit the individual media manually so the reproducing times of the corresponding parts of the media coincided to each other. However, because of the complex relationship between the times of different media, it was difficult work for the editor.
In the Japanese Patent Provisional Publication No. Hei 5-3561, a method to produce a multimedia representation is disclosed which makes the reproducing times of image and sound coincide, when they are different, by extending or reducing the times of image and sound. However, this method extends or reduces the reproducing time of each media to the mean values of the media without any consideration for the characteristics of the sound and images. Sometimes this results in unnatural or incomplete reproduction, especially when the media whose reproducing time is extended or reduced is an image. For example, a scene of a man walking too fast or too slow will be notably unnatural.